Skip to main content

Statistical independence for the evaluation of classifier-based diagnosis